Fractals, ZZ & the Pissing Dog | Page 17 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

ECN && STP && Scalping && Hedging - Broker Review - Oct 2009 24 replies

Alert on Fractals (MT) 2 replies

Zigzag & Fractals 10 replies

EA wanted for sound alert depending on fractals condition 2 replies

Fractals Edge Fx 3 replies

Attachments: Fractals, ZZ & the Pissing Dog
Exit Attachments

Fractals, ZZ & the Pissing Dog

  • #321
  • Jan 25, 2014 2:53am Jan 25, 2014 2:53am
Quoting gg53
{quote} emonts, I'm still here, not going anywhere... I said from day 1 that I'm also working and testing Renko/Range bars. It's exactly the same system and algorithm - but I think it will bring better yield with some innovative R/R & M/M. G.
Clear and thanks.

BR.
Huub
It is not important how often you fall down, but how often you get up again
  • #322
  • Jan 25, 2014 2:59am Jan 25, 2014 2:59am
good luck and thanks for all your effort GG.you are inspire for a total new 'noob' trader like my self.
  • #323
  • Jan 25, 2014 3:12am Jan 25, 2014 3:12am
Quoting jmn5611
gg53 Thanks for starting a TE. It really makes you legit. Now matter what happens going forward. Everyone can see it. That alone brings honor to what you are trying to do. Serious people look at whether or not you are willing to expose your trading in real time, with a neutral arbiter.
jmn5611,

"Expose my trading" - that's not going to happen here.
The purpose is to test a theory that knowing what to trade, a little "edge", R/R, M/M combined together can achieve a lot.
It's not going to recommend trading in such a way on regular basis.

"Serious people" .... sounds scary... ;-}

G.
  • #324
  • Edited 5:25am Jan 25, 2014 3:57am | Edited 5:25am
hello gg. here's a brief outline on how i would trade it:

1. entry/exit: signal to signal, so always in the market.

2. size: 1.0 lot/ per each $1000 on account

3. daily target: net 30 pips, which represent 30% of account

4. agree with FXEZ, no decrease in size after a loss, increase after each win

5. add positions using renko wicked continuation bars within swing, with a scalping tp (5-10 pips, depending on brick size). adjust size after each win, always with a ratio of 1.0/$1000

studying your 5mn screenshots, this is definetely achievable, and it would translate in doubling account every week. trade size is what i use for my renko fractal breakout strategy.

hope it helps.
  • #325
  • Jan 25, 2014 5:11am Jan 25, 2014 5:11am
@ gg53
Very good , and I wish you good luck with your test.

I have some questions about " errors predictions " that i would ask for to understand a little more of your system .

There are two types of prediction errors :

1 - one ZZ appears and our signal is at a distance of 3 candles . We expect confirmation ZZ ( 5 bars ) .
Fails . We started to calculate from bar 5 after ZZ confirmed .
We store ZZ confirmed as follows :

- We store in our database fractal sequence
- We store in our database length ( -1 bar)
- We filter with SET1 and store the sequence in Set3 bars.

2 - The fractal prediction is not expected ( Monty Hall) . For example :
a) We have the fractal pattern
[1 0] = 273 Occurs

and we predict :
[ 1 0 1 ] = 793 Occurs
[1 0 0] = 211 Occurs

The pattern [ 1 0 1 ] has higher probabilities .
We wait , but the market gives us the pattern [1 0 0].
Here there is a prediction error that we must correct similar looking pattern.

b ) But as we meet in the following case ?
We have the fractal pattern [ 1 0 1 ] = 793 Occurs
And we predict :
[ 1 0 1 1] = 213 Occurs
[ 1 0 1 0] = 278 Occurs

But in the end , do not get signal in the pattern [ 1 0 1 ] and were shown
[ 1 0 1 0].
Here we also have a prediction error . We apply Monty Hall?
I 'm on the right way ?

thank you very much
  • #326
  • Jan 25, 2014 5:45am Jan 25, 2014 5:45am
Quoting burnssss
@ gg53 Very good , and I wish you good luck with your test. I have some questions about " errors predictions " that i would ask for to understand a little more of your system . There are two types of prediction errors : 1 - one ZZ appears and our signal is at a distance of 3 candles . We expect confirmation ZZ ( 5 bars ) . Fails . We started to calculate from bar 5 after ZZ confirmed . We store ZZ confirmed as follows : - We store in our database fractal sequence - We store in our database length ( -1 bar) - We filter with SET1 and store the sequence...
Hi burnsss,

I think GG gonna post it if he get a little time to make an explanation for us. This part is missing yet:
Handling Prediction errors:- Coming soon...
So I think better to wait for it.

GG,

good luck for your trial, I think with your system your monthly target is absolutely achievable!
Hope you keep support us I can not wait to "finish" the algo and after test it on renko charts too. I like renko very much.

regards
  • #327
  • Jan 25, 2014 5:58am Jan 25, 2014 5:58am
Quoting burnssss
@ gg53 Very good , and I wish you good luck with your test. I have some questions about " errors predictions " that i would ask for to understand a little more of your system . There are two types of prediction errors : 1 - one ZZ appears and our signal is at a distance of 3 candles . We expect confirmation ZZ ( 5 bars ) . Fails . We started to calculate from bar 5 after ZZ confirmed . We store ZZ confirmed as follows : - We store in our database fractal sequence - We store in our database length ( -1 bar) - We filter with SET1 and store the sequence...
burnssss,

1. Yes, but I'm still looking for a better way than just "more than 3 bar away". Sometimes the "error" is a better entry for trading.
2. Yes, good description, but there are more cases when this kind of "errors" occur - and we have to change our predicted "door".

G.
  • #328
  • Jan 25, 2014 6:06am Jan 25, 2014 6:06am
Quoting gg53
{quote} burnssss, 1. Yes, but I'm still looking for a better way than just "more than 3 bar away". Sometimes the "error" is a better entry for trading. 2. Yes, good description, but there are more cases when this kind of "errors" occur - and we have to change our predicted "door". G.
@ gg53
Thank you very much for your reply.
1 - You could evaluate this signal as a function of ...... That would be the possible filter, considering the range you've been using since the beginning (3 candles) is not a bad decision.
2 - What more cases of error messages?

This system has a lot to do with similarity. It tries to find a balance between error probability and signals.
I begin to understand more.

Many thanks
  • #329
  • Jan 25, 2014 6:20am Jan 25, 2014 6:20am
Quoting burnssss
{quote} @ gg53 Thank you very much for your reply. 1 - You could evaluate this signal as a function of ...... That would be the possible filter, considering the range you've been using since the beginning (3 candles) is not a bad decision. 2 - What more cases of error messages? This system has a lot to do with similarity. It tries to find a balance between error probability and signals. I begin to understand more. Many thanks
burnssss,

I'll explain this issue in a separate post, but one more case is when everything "fits" and you are waiting for "next bar" and instead you are getting a fractal.
I fail to see the similarity to the similarity... ;-}

G.
  • #330
  • Jan 25, 2014 6:23am Jan 25, 2014 6:23am
Quoting gg53
{quote} burnssss, I'll explain this issue in a separate post, but one more case is when everything "fits" and you are waiting for "next bar" and instead you are getting a fractal. G.
Sorry gg53,
I'll wait for your post.
We have 3 types of errors for problem 3 doors lol.
Many thanks
  • #331
  • Jan 25, 2014 10:38am Jan 25, 2014 10:38am
Quoting gg53
{quote} FXEZ, You mentioned average DrawDown - based on what strategy? I don't think that implementing that indicator on Renko charts can be called "reversal system". Account size is $10,000 and needs to be doubled within roughly a month. That's ~20 trading days, ~$500 a day profit. Calculations and statistics start from that point. G.
Hi G,

It sounds like you're just conceiving of how to trade the indicator as a system. I think of what size to trade as a sort of equation that can be solved, given the individual's goals. If you are trying to solve the "double your account in 1 month" problem, you need to know how many total pips on average the strategy returns as a minimum, and preferably also system drawdown so you can have a general idea of how much leverage needs to be used to achieve that goal or where too much leverage will lead to undesired results.

What I do is drop in some code into an indicator that will give a rough idea that if a given indicator were turned into an always-in reversal strategy what are the total pips of profit, win%, avg profit per trade and drawdown. Given those numbers I can calculate the average weekly profit in pips. Given that, it is fairly straightforward to create a spreadsheet that will show profit progression from a starting figure for a given # of pips and a chosen leverage amount. It is also fairly easy to create variations on the always-in strategy by adding conditions once you have an idea of the base performance numbers.

I have some indicator drop-in code I can share with you if you are interested.

Patrick
  • #332
  • Jan 25, 2014 10:44am Jan 25, 2014 10:44am
Quoting gg53
{quote} jmn5611, "Expose my trading" - that's not going to happen here. The purpose is to test a theory that knowing what to trade, a little "edge", R/R, M/M combined together can achieve a lot. It's not going to recommend trading in such a way on regular basis. "Serious people" .... sounds scary... ;-} G.

Well, data is always useful, and it is good to see what your inputs can do.
If you are good at something, never do it for free--Joker
  • #333
  • Edited 12:43pm Jan 25, 2014 12:33pm | Edited 12:43pm
Quoting FXEZ
{quote} Hi G, It sounds like you're just conceiving of how to trade the indicator as a system. I think of what size to trade as a sort of equation that can be solved, given the individual's goals. If you are trying to solve the "double your account in 1 month" problem, you need to know how many total pips on average the strategy returns as a minimum, and preferably also system drawdown so you can have a general idea of how much leverage needs to be used to achieve that goal or where too much leverage will lead to undesired results. What I do is...
FXEZ,

It's not that simple...
The strategy or system is not "always in". It selects what it thinks going to be the best trade from various currency pairs.
The selection is based on Currency strength/wekness, potential yield, and only than it implements the "ForexGT_Arrow", if signaled.
The first part is the "What" to trade - potential best use of the fund.

Limitations are: no more than 3-4 trades at one time, no more than 70% of total account size.

G.
  • #334
  • Jan 25, 2014 9:16pm Jan 25, 2014 9:16pm
Quoting gg53
{quote} FXEZ, It's not that simple... The strategy or system is not "always in". It selects what it thinks going to be the best trade from various currency pairs. The selection is based on Currency strength/wekness, potential yield, and only than it implements the "ForexGT_Arrow", if signaled. The first part is the "What" to trade - potential best use of the fund. Limitations are: no more than 3-4 trades at one time, no more than 70% of total account size. G.
Please explain GG. How do you quantify potential yield algorithmically?
  • #335
  • Jan 25, 2014 10:45pm Jan 25, 2014 10:45pm
Quoting gg53
{quote} Can you roughly double an account ? Trade-explorer is set up and waiting to start on next week open. Strategy: Little "edge" applied on Renko charts, statistical tools and "Out-of-the-Box" thinking on Risk/Reward & Money-Management. Wish me luck... G.

Wish you luck gg!! with the 5 min charts you have shown about this indicator, you should have little difficulty in reaching your goal.

however as FXEZ said, the goal of "double in 1 month" is just a function of leverage. the underlying trading method's quality is more directly measured by average monthly profit/max drawdown. once you have that # as high as possible, after that it is just a calculation to know what leverage to use to achieve what goal (and also know what is feasible or not).

best funds have a long term ratio of 3:1 of yearly profit : max drawdown (think of SAC, MLP etc).. Seems like your are better than that :-)
Think : Trade : Live Life
  • #336
  • Edited 12:51am Jan 26, 2014 12:31am | Edited 12:51am
Quoting Thoughts
{quote} Please explain GG. How do you quantify potential yield algorithmically?
Thoughts,

Simple.... Historical database of currency pairs.
The questions (for each pair):
What is the average pips when you get a signal at certain time of day?
How it relates to currency pair daily movement (pips/day)?
How much the currency pair done so far?
How long does it take to win those pips?
What is the DrawDown?
What is the Win/Loss rate?
What is the CURRENT individual currency strength relation?

Each one of those params have its coefficient. Currency strength relation have the highest coefficient.

Below a certain number - don't trade that signal.
If more than one signal - trade the highest ones, in descending order.
Lot sizing according to my variation of Kelly & Optimal f multiplied by the above coefficient.

The historical database is devided into days, so each query relates to data within a certain day of week.

G.
  • #337
  • Jan 26, 2014 1:07am Jan 26, 2014 1:07am
Quoting stt
{quote} Wish you luck gg!! with the 5 min charts you have shown about this indicator, you should have little difficulty in reaching your goal. however as FXEZ said, the goal of "double in 1 month" is just a function of leverage. the underlying trading method's quality is more directly measured by average monthly profit/max drawdown. once you have that # as high as possible, after that it is just a calculation to know what leverage to use to achieve what goal (and also know what is feasible or not). best funds have a long term ratio of 3:1 of yearly...
stt,

I don't think it's just leverage, although it's big part of it.
It's also best usage of your funds - i.e WHAT to trade for max results.
In any case, I doubt that 5Min charts will be effective in reaching that target.
I think that higher TF's are required to achieve that target.

G.
  • #338
  • Jan 26, 2014 6:10am Jan 26, 2014 6:10am
Quoting gg53
{quote} stt, I don't think it's just leverage, although it's big part of it. It's also best usage of your funds - i.e WHAT to trade for max results. In any case, I doubt that 5Min charts will be effective in reaching that target. I think that higher TF's are required to achieve that target. G.
But with 5 minute charts, you can have tighter stops.

Maybe if you moved to tick charts, you could have tight stops and large take profits.

edit: oh yeah, maybe that is why you developed it with the new renko charts.
  • #339
  • Jan 26, 2014 10:04am Jan 26, 2014 10:04am
Quoting Starry
{quote} But with 5 minute charts, you can have tighter stops. Maybe if you moved to tick charts, you could have tight stops and large take profits. edit: oh yeah, maybe that is why you developed it with the new renko charts.
Starry,

Spot on!

G.
  • #340
  • Jan 27, 2014 12:45pm Jan 27, 2014 12:45pm
Quoting gg53
{quote} Thoughts, Simple.... Historical database of currency pairs. The questions (for each pair): What is the average pips when you get a signal at certain time of day? How it relates to currency pair daily movement (pips/day)? How much the currency pair done so far? How long does it take to win those pips? What is the DrawDown? What is the Win/Loss rate? What is the CURRENT individual currency strength relation? Each one of those params have its coefficient. Currency strength relation have the highest coefficient. Below a certain number - don't...
Thanks for explaining GG..your statistical approach towards RR sounds similar (in principle) to something I read in one of Victor Sperandeo's books..but I think he was looking at much longer time frames. I'm curious though, why not just trade 'always in' with the indicator?
0 traders viewing now